Lesson 3.6Lesson 18 of 30

Claude Code in CI/CD

Integrate Claude Code into CI/CD pipelines

Key points

1In CI, run claude -p "..." (--print): Claude Code answers non-interactively and exits. Without -p it opens an interactive session and the job hangs waiting for input.
2For findings a script must parse, add --output-format json with --json-schema: each finding comes back as validated fields (file, line, severity, message) that the job posts as inline PR comments.
3CI-invoked Claude Code reads the committed project CLAUDE.md. Put testing standards, what makes a test valuable, fixture conventions and review criteria there, not in a personal file or a long prompt in the workflow YAML.
4Review with an independent instance: the session that generated the code keeps its own reasoning and is worse at spotting its mistakes. Run the review as a separate claude -p call, not a resumed generator session.
5Per-run context goes in the prompt: on re-review after new commits, pass the prior findings and ask for only new or still-unaddressed issues; for test generation, pass the existing test files so covered scenarios are not duplicated.

What the exam guide checks here

Objective wording from the exam guide (English).

3.6.K1-p / --print runs Claude Code non-interactively
3.6.K2--output-format json and --json-schema enforce structured CI output
3.6.K3CLAUDE.md provides project context to CI-invoked Claude Code
3.6.K4Session isolation: the generating session reviews its own code worse than an independent instance
3.6.S1Use -p in CI to avoid interactive hangs
3.6.S2Use --output-format json with --json-schema for machine-parseable findings posted as PR comments
3.6.S3On re-review include prior findings and report only new or unaddressed issues
3.6.S4Provide existing tests so generation avoids duplicate scenarios
3.6.S5Document testing standards, valuable test criteria and fixtures in CLAUDE.md

Answer scenario questions on this topic. Every option comes with an explanation.